Circuit help: switching a load only at peak of mains

[Phil Allison]

"JB"

This is really only to test the absolute worst case for inrush current for
the given load. I like to have some 'headroom' when I'm specifying the
relays in our products which switch these loads in the field, so I need
worst case scenarios. It only has to work on 50Hz supplies and is only for
in house testing purposes here in the lab.

** You should seriously consider changing from relays to zero crossing triac
control. This can work well with electronic LED, incandescent and small
transformer loads used with low voltage halogen.

In the case of LED and CFL lamps, the high inrush surges commonly seen are
reduced by a factor of at least 5 because voltage rise time is controlled at
0.1V/uS with zero switching.

The MOC3062, 240V zero crossing opto driver and a 16A triac ( possibly a
snubberless type ) would be a good start.


.... Phil

 

[Lasse Langwadt Christensen]

Den onsdag den 18. september 2013 10.39.26 UTC+2 skrev Phil Allison:

"JB"











** You should seriously consider changing from relays to zero crossing triac

control. This can work well with electronic LED, incandescent and small

transformer loads used with low voltage halogen.



In the case of LED and CFL lamps, the high inrush surges commonly seen are

reduced by a factor of at least 5 because voltage rise time is controlled at

0.1V/uS with zero switching.



The MOC3062, 240V zero crossing opto driver and a 16A triac ( possibly a

snubberless type ) would be a good start.

all in one:
http://sharp-world.com/products/device/lineup/data/pdf/datasheet/s116s02_e.pdf

-Lasse

 

[Phil Allison]

"Lasse Langwadt Christensen"

Phil Allison:

all in one:
http://sharp-world.com/products/device/lineup/data/pdf/datasheet/s116s02_e.pdf

** Bad idea for a small manufacturer.

Single source, relatively expensive and liable to become unavailable.

Suitable 6 pins optos and 3 pin triacs are not going away any time soon.



.... Phil

 

[George Herold]

Differentiation picks up all kinds of line noise.

Ahh very good point. Thanks Jim.
I suppose you could put two low passes in the circuit (with some sort of tweak) to get more or less than a 90 degree phase shift. But at that point you might as well add an opamp and make an all-pass.

This blank line insertion is becoming profoundly annoying.

You are absolutely correct! It's now become down right *rude* to use google groups for usenet. (Or a major pita to remove all the extra lines.) I'll have to write them a note, and see if eternal september will take me back..
(Maybe tomorrow :^)

George H.

 

[JB]

---
I see something with a push-button switch which enables a
zero-crossing detector which starts a 5 millisecond delay timer which
sets an R-S latch when it times out, which turns on a TRIAC at 90
degrees and holds it on until the R-S latch is reset by a push-button
switch.

Would that work for you?

This is sort of what I had in mind. It seems the simplest and we can
implement this easily here.
Thanks to all for suggestions.

JB

 

[Jasen Betts]

Shit, 12.5.......

I'm changing my handle to mushbrain, hey I just woke up.

Any advance on 12.5?

who has 20Hz power?

 

[Anthony Stewart]

The simple way using a bridge between AC and Load is to put a small series resistor on the ground side of a large cap with some load for small ripple.. This ensures the cap current charges only for a short time before peak when you can use the cap current sense to drive a triac with ~1V just before peak voltage.

It is useful to know that random switches such as Relay contacts without ZCS rely on the random phase for calculating the cummulative aging effects ofsurges. to rate lifetime. This is how Omron, the experts in Relay, design factors lifetime vs load current and number of cycles. This means you don't need to design a tester for this and can calculate the worst case inrush based on the ESR of the load cap.


Arc currents are worst for contacts on Caps loads when closing and on inductive loads when opening.

Snubbers reduce the harmful effects of inductive loads and ICL or NTC inrush current limiters reduce the harmful effects of low ESR capacitive loads when discharged and switched at peak voltage.

If need references search the Omron site or just ask me. This random phase on AC loads are the soe reason why Relay contacts have higher current ratings than DC loads as the AC switch current is less than peak worst case most of the time.

Whereas in DC it is always worst case with 0V for initial conditions.

 

[Phil Allison]

"Anthony Stewart"

The simple way using a bridge between AC and Load is to put a small series
resistor on the ground side of a large cap with some load for small ripple.
This ensures the cap current charges only for a short time before peak when
you can use the cap current sense to drive a triac with ~1V just before peak
voltage.

** Totally misses the point of the OP's question.

" I have a series of measurements to make of the inrush current ... "

FYI: " inrush current " refers to the one time transient at initial switch
on.


ZCS rely on the random phase for calculating the cummulative aging effects
of surges. to rate lifetime. This is how Omron, the experts in Relay,
design factors lifetime vs load current and number of cycles. This means
you don't need to design a tester for this and can calculate the worst case
inrush based on the ESR of the load cap.

** Totally off with the fairies.

" ....the inrush current of a range of LED drivers and electronic ballasts"

The OP is not designing the loads, they are out there in the market already.

The worst examples of which are yet to appear.


.... Phil

 

[Anthony Stewart]

Phil you failed to comprehend.

A peak crossing Triac is simple taking to 90 phase shift in a low pass filter as I explained sensing cap current at peak.

Another way is to use a 5ms one shot after each zero crossing.

Use a small signal diode bridge to make a full wave rect signal. Use 100:1resistor divider to get down to logic levels or use an opto coupler if unsure. A full bridge is equal to an XOR gate that will give you a 100Hz rate Zero Crossing Switch pulse from 50Hz in. Usinga 5 ms one shot delay to fire another one shot to trigger the Triac will guarantee peak voltage switching.

Since Phil failed to comprehend the accumulated wear factor of a random switch on AC loads and how surge varies dependant on phase which affects surge, Relay engineers rely on this random nature to increase the rating lifetime of a Relay. This would not apply to dimmer Triacs as the repetition rate is at line frequency and non-dimmable reactive ballasts would fry the Triac in a few seconds hours or days depending on reactance of CFL or LED ballast.

In Omron's site they will specify conditions for reactive loads such as,


Note:If the L/R of an inductive load exceeds 7 ms with a Model for a DC Load, the arc interruption time must be less than approximately 50 ms to use the Relay. Design the circuit so that the arc interruption time is 50 ms or less. *These values apply to a switching frequency of 30 times per minute.


In Capactive loads, the critical factor is the current limiting resistance of the switch, diodes and the Cap ESR.

I stated if you know the reactance and the ESR, you can save the trouble oftesting a product, with surge currents at peak voltage, just measure the input impedance . But if you don't know, then by all means test it.

 

[Phil Allison]

"Anthony Stewart"

Phil you failed to comprehend.

** Like hell I did.

Listen pal - YOU are a raving nut case, obsessed with your mole hill of
knowledge on an irrelevance to the OP's problem.

PLUS you are breaking all the rule of usenet posting.

Snipping the entire context and posting speeches is ABSURD !!



..... Phil

 

[Anthony Stewart]

Phil,

TO be prudent, ethical, practical and rational please stick to the topic and suppress your personal remarks. Use intelligent corrections if you dispute my assertions.

1. 90 Phase shifter to trigger Triac at peak voltage using RC current sense.. Works.

2. Relays with AC contact ratings are rated higher than DC because DC has max surge each time, AC does not due to random phase and surge currents are thus averaged to extend peak current ratings.

3. AC also suppresses sustained arcs from zero cross on turn off, but question here refers to turn on surge.

4. Zero crossing clock derived from Line AC sinewave with frequency doubler( bridge or XOR gate followed by 5 ms one shot delay for 90 deg phase shift at 50Hz will be provide trailing edge precision to trigger a 2nd one shotto fire and latch a Relay for RC time constant of one shot or simply fire a Triac for next cycle depending on desired component to be tested with desired loads.

5. No one knows more about Relays than Omron

If you disagree be sure to back it up with facts and not weightless opinions.

 

[Phil Allison]

"Anthony Stewart PISS OFF "


Listen pal - YOU are a raving nut case, obsessed with your mole hill of
knowledge on an irrelevance to the OP's problem.

PLUS you are breaking all the rule of usenet posting.

Snipping the entire context and posting speeches is ABSURD !!

So **** OFF !!!!!!!!!!!


..... Phil

 

[Phil Allison]

"John Fields"

---
Relay contacts rated for a particular AC current and voltage will
carry the same DC current but must be operated with a lower voltage in
order to quench the arc which will occur when the contacts open. That
arc will automatically extinguish when the voltage across the AC
contacts crosses zero volts
---

** 240VAC rated relays normally have a DC rating with the same current value
but at no more than 24V - use them much above that voltage and an arc will
form at switch off that never extinguishes.

240VAC rated circuit breakers are typically rated at 50V DC, but they have
contacts that open much faster and wider than most relays.


.... Phil

 

[Anthony Stewart]

I recognize your illness. Bite your tongue if wish to stay around

 

[Phil Allison]

"Anthony Stewart PISS OFF "


Listen pal - YOU are a raving nut case, obsessed with your mole hill of
knowledge on an irrelevance to the OP's problem.

PLUS you are breaking all the rule of usenet posting.

Snipping the entire context and posting STUPID speeches is ABSURD !!

So **** OFF !!!!!!!!!!!


..... Phil

 

[Anthony Stewart]

To back up my assertion that no one knows more about Relays than OMRON , I offer this to anyone who wants to learn more about Relays, drivers etc.

http://www.omron.com/ecb/products/pdf/precautions_pcb.pdf

Example of a well written Relay Spec , http://teachmetomake.files.wordpress..com/2011/02/relay-g5v1.pdf with Endurance vs contact current for AC vs DC ..

Note it is limited by power dissipation for small size.

Vacuum-sealed Relays? Although air can sustain arc burning it cannot prevent it. Vacuum sealed Relays are not used in small items due to poor heat transfer of contacts which is the limiting factor for contact ratings. But they make many Hermetic sealed Relays for explosion-prevention environmentswith UL safety rating.

Although this goes beyond, OP's question, it may shed light on proper use and how to choose and specify all aspects of Relays.

If you know how XOR gates work when output goes high if only one input is high, you will notice a Diode Bridge performs the same task, when the outputDC goes high only if one or the other ~ input is higher than the other. this fact allows both devices to be used for frequency doubling so that pulses can be created for + &- transitions to make a positive pulse out. With XOR you only need to delay a clock input with small RC delay to produce positive ZERO crosssing pulses. similar with Bridge and RC current sense I suggested, small delay will be 0 degree on leading edge , and max delay of RC filter is 90 deg on trailing edge of diode to cap charging current. If threshold to Triac is chosen properly this can work well enough to trigger at near peak of 90 shifted voltage.

A picture would be clearer. But my comments are only gobbly-gook to those who cannot understand.

 

[rickman]

Shit, 12.5.......

I'm changing my handle to mushbrain, hey I just woke up.

I assume you are in the UK, but it would be 5 ms, no? 20 ms is a full
cycle and the delay needs to be 1/4 cycle or 5 ms. 4.167 ms when the
line frequency is 60 Hz.